The mistreatment of company-stock options is beginning to grab some headlines, and for good reason. Which is not to say that options are inherently bad — to the contrary, if you're receiving them, they're quite wonderful. Stock options are capitalism's most popular 'heads you win, tails you break-even' proposition. As an employee or, in particular, as an executive of a company, you can benefit from a rise in your company's stock price without risking any loss from its decline.

As a shareholder in a company, however, the equation isn't so favorable. For one thing, if the stock you own declines in price, you lose money; for another, those options given to employees and execs are worth something, and that something comes out of your pocket. Of course, you might not realize this by reading a company's income statement since option accounting is one of the biggest gaps in GAAP (Generally Accepted Accounting Principles).

Companies are required to account for the impact of options, but in a manner which doesn't much help investors. For example, in fiscal 2000, Cisco issued 295 million options to its employees (8 million to its top six execs) which represented about 4% of the shares outstanding. On its income statement, Cisco accounted for the additional shares (which will be issued when the options are exercized) by increasing the number of "fully-diluted" shares — earnings per share were reduced accordingly.

Sounds fair? Not really. The problem is that these options, when properly valued, reduced earnings by a great deal more than 4%. Based on the widely used Black-Scholes model, the value of the options granted were $5 billion, more than the entire pre-tax earnings of $4.3 billion!

Why, then, doesn't the FASB require more accurate accounting? For one thing, GAAP accounting uses rules which try to maintain consistency and objectivity in financial disclosures. Black-Scholes is consistent; as for its objectivity, that could be argued either way. The fact that its valuations are inaccurate, based as they are on uncertain future events, doesn't undermine its usefulness — GAAP accounting is already inaccurate (based ironically on its reliance on historical valuations for assets).

The current accounting for options is a bit like the fellow who dropped his wallet in an alley but was looking for it under a distant street lamp because the light was better. And wasn't it no less an authority than John Maynard Keynes who said that he'd "rather be vaguely right than precisely wrong"?

If the Black-Scholes valuation is imperfect, it nonetheless has merit, particularly when used in conjunction with an estimation of the replacement cost of options. This approach, suggested by Martin Hutchinson, simply relies on determining the cost to a company of buying back the same number of options which are exercized. In Cisco's case, 176 million options were exercized in fiscal 2000 at an average price of $5.75 per share, or $1 billion. To buy-back these shares, and keep shareholders from being diluted, the cost would have approximated $9 billion — net cost, $8 billion!

Again, the question is, why don't companies have to account for the cost of options in their income statements? A partial answer is that they're now required to disclose the potential impacts in the footnotes to their annual reports. Aside from these disclosures being less than obvious to most investors, there is also a question of methodology.

Cisco's footnotes indicate that, when accounting for the impact of options "based on specified valuation techniques that produce estimated compensation charges" as required by SFAS 123 ("Accounting for Stock-Based Compensation"), earnings are reduced by 40%. This is a meaningful hit, but still quite a bit less than the impact of the two approaches suggested above, which more than wiped out earnings for the year. The difference here is one of timing: SFAS 123 amortizes the costs of options (over their vesting periods); the other two approaches expense the costs in the year they're incurred.

At Microsoft, the cost for fiscal 2000 using the Black-Scholes model was more than $9 billion and the replacement cost was almost $20 billion — this for a company whose pre-tax earnings were $14.3 billion. Under 123, by contrast, the charge was less than $2 billion. (To its credit, Microsoft is one of the rare companies which highlight the SFAS 123 costs in its quarterly financial releases.) At Dell Computer, the cost in fiscal 2000 using Black-Scholes was more than $700 million and replacement cost was almost $3 billion vs. pre-tax earnings of $2.5 billion. Under SFAS 123, the charge would have been $329 million.

Returning to the underlying question, perhaps the best reason why this country's public companies disclose as they do is the obvious reason. As The Economist noted in its Jan. 27 issue, "The main argument advanced for not putting options into the profit-and-loss account is perfectly simple: to do so would cost American companies too much."

Now, if only companies could exclude depreciation charges from the income statement...what a boost to earnings!

Suggestion: Go to Optionville — Where all CEOs are above-average